Syncsort Sorting out Big Data
Syncsort’s Lone Jaffe, new CEO of Data Integration1 as of June 2013, has set his company’s sights on owning Big Data sorting. The origin of Syncsort’s prowess in processing data quickly comes from a protracted fight with IBM’s mainframe sort program in the 1970s. In a world where batch tape and disk sorts times determined the length of the batch window, which in turn determined the availability hours of online systems, IBM could not shake them off. Syncsort emerged with the world’s best collection of large-scale sorting, merging, copying, and other data management utilities, together with the ability to programmatically enable operation stability.
The world is moving towards these skills again. One Big Data key business value driver is the enablement of near real-time data analytics from many data sources, and the ability to operationalize these processes to automate change in real-time online systems. Sorting is a large component of elapsed time in both the ETL and selection processes. Jaffe is betting that the same technologies (with the addition of large in-memory sorts) are going to be needed for Big Data.
His first acquisition is the UK-based Circle Computer Group. Circle’s flagship product, DL/2, provides a software engine that enables the transparent migration of applications accessing data held in IBM’s mainframe hierarchical IMS/DB database to IBM’s relational DB2, without requiring any application changes. The same process can be applied to VSAM-based applications with VS/2, and IBM IMS Fast Path is on the roadmap.
Mainframe data in general is not as voluminous as other enterprise data but has high data change rates of high enterprise value. One important value of this transformation from hierarchical to relational is that the data can now become part of a real-time Hadoop ETL process including Syncsort’s DMExpress, and is a fast and easy way to integrate mainframe data into real-time analytic processes. Cloudera and Syncsort recently announced a technology partnership focused on enabling enterprises to leverage mainframe data in Hadoop. Earlier this month, Syncsort announced a new partnership with Tableau Software.
Action Item: CIOs and CTOs should be setting out a roadmap for the introduction of near real-time data analytics, and partnering with companies that can provide the key technologies to enable automation of rapid change to business systems. Syncsort is likely to be one such important technology to reduce the elapsed time of complex data transformation, especially for companies that hold large amounts of important data in IBM mainframe hierarchical databases.